Ovarian cancer causes more deaths than any other female tract reproductive malignancy. Five-year survival has remained relatively unchanged for the past three decades, and drug resistance remains a major therapeutic barrier. During the initial funding period, we demonstrated that increased DNA methylation and alterations in histone methylation are strongly associated with ovarian cancer chemoresistance. A prevalent model of chemoresistance, and carcinogenesis in general, is based on tumor growth due to the behavior of abnormal stem cells, and it is hypothesized that the inadequacy of standard ovarian cancer therapies is largely due to their failure to target ovarian tumor stem cells (resulting in inevitable relapse). A critical role for repressive epigenetic (noncoding) modifications has become increasingly evident for the establishment and maintenance of both normal and cancer stem/initiating cells, and the long-term goal of our laboratory is to elucidate the role of epigenetics in ovarian tumor formation and progression. By exploiting recently described characteristics of cancer stem cells, we have now isolated and substantiated ovarian cancer-initiating cells from human ovarian tumors. The overall hypothesis to be tested is that epigenetic events in ovarian cancer- initiating cells govern tumor drug resistance and response to chemotherapy. While epigenetic modifications likely affect multiple pathways in cancer stem cells, including self-renewal, proliferation and differentiation, we will specifically examine the epigenetic events in ovarian cancer-initiating cells that contribute to drug resistance; reversal of such events may allow for the "epigenetic resensitization" of chemoresistant tumor stem cells. The goal of Aim 1 is to identify DNA "methylation signatures" associated with chemoresistance in ovarian cancer-initiating cells. We will further examine 40 methylated genes previously identified by our group as associated with disease-free survival, using genome-wide profiling of DNA methylation. In Aim 2, a high- throughput epigenetic analysis, chromatin immunoprecipitation-to-microarray hybridization, will be used to identify genes possessing specific activating and repressive histone marks in ovarian cancer-initiating cells. Integrating this histone modification data with the DNA methylation data from Aim 1 will facilitate the identification of pathways responsible for ovarian cancer-initiating cell chemoresistance. The direct targeting of ovarian cancer-initiating cells, using epigenetic and conventional agents, is the objective of Aim 3. We hypothesize that inhibitors of DNA methylation and histone modifications can disrupt epigenetic repressive modifications necessary for tumorigenic differentiation of the ovarian cancer-initiating cells responsible for propagating an entire tumor, allowing for their direct therapeutic targeting. Knowledge of the epigenetic events present in ovarian cancer-initiating cells will result in a more complete understanding of this devastating disease and will also provide valuable insight into new therapeutic targets. This proposal is highly responsive to PA-05-086, based on its overall objective to establish ovarian cancer stem cell epigenotypes, develop a comprehensive model of altered pathways responsible for chemoresistance, and examine the possible direct disruption of the activity of these cells by epigenetic therapies. PUBLIC HEALTH RELEVANCE: Of all female cancers, ovarian cancer is by far the deadliest. A new opportunity in ovarian cancer research is that ovarian cancer is believed to contain a small population of "cancer stem cells" responsible for growing an entire tumor. While conventional cancer chemotherapies kill most of the cells in a tumor, they may miss these cancer stem cells, allowing the tumor to re-grow. "Epigenetic" changes (changes in DNA structure) may be a way to characterize ovarian cancer stem cells, and new drugs that alter those DNA structural changes may represent a new type of therapy against those cells, allowing complete killing of the tumor. [unreadable] [unreadable] [unreadable]